Mengjio Jing, Hongf Xu, Ninfu Yng, Gnghu Li, Ynfeng Ding,b,,Mtthew J.Pul, Zhenghui Liu,b,

aCollege of Agriculture,Nanjing Agricultural University,Nanjing 210095,Jiangsu,China

bCollaborative Innovation Center for Modern Crop Production,Nanjing Agricultural University, Nanjing 210095,Jiangsu,China

cPlant Science,Rothamsted Research,Harpenden,Hertfordshire AL5 2JQ,UK

Keywords:

ABSTRACT Carbon isotope composition (δ13C) of a plant organ is an inherent signature reflecting its physiological property,and thus is used as an integrative index in crop breeding.It is also a non-intrusive method for quantifying the relative contribution of different source organs to grain filling in cereals.Using the samples collected from two-year field and pot experiments with two nitrogen (N) fertilization treatments, we investigated the temporal and spatial variations of δ13C in source organs of leaf, sheath, internode, and bracts, and in sink organ grain.Constitutive nature of δ13C was uncovered, with an order of leaf (?27.84‰) ＜grain(?27.82‰) ＜sheath (?27.24‰) ＜bracts (?26.81‰) ＜internode (?25.67‰).For different positions of individual organs within the plant, δ13C of the leaf and sheath presented a diminishing trend from the top (flag leaf and its sheath) to the bottom (the last leaf in reverse order and its sheath).No obvious pattern was found for the internode.For temporal variations, δ13C of the leaf and sheath had a peak (the most negative) at 10 days after anthesis (DAA), whereas that of the bracts showed a marked increase at the time point of anthesis,implying a transformation from sink to source organ.By comparing the δ13C in its natural abundance in the water-soluble fractions of the sheath,internode,and bracts with the δ13C in mature grains, the relative contribution of these organs to grain filling was assessed.With reference to the leaf, the internode accounted for as high as 32.64% and 42.56% at 10 DAA and 20 DAA, respectively.Meanwhile, bracts presented a larger contribution than the internode, with superior bracts being higher than inferior bracts.In addition, N topdressing reduced the contribution of the internode and bracts.Our findings clearly proved the actual significance of non-foliar organs of the internode and bracts for rice yield formation, thus extending our basic knowledge of source and sink relations.

1.Introduction

Increasing rice yield is undoubtedly important since there are 821 million people regularly experiencing hunger in the world[1].However, it is not trivial for rice improvement to keep up with the rate of population growth and consumer demand[2].The formation of grain yield is,rather,a multifaceted process governed by a large array of genes and affected by variable environmental factors.Crop geneticists and breeders have adopted a set of strategies ranging from marker assisted selection on target genes to genomic selection on genes and their regulatory networks, with the aim to enhance the efficiency of improving complex agronomical traits [3].In this regard,a more complete understanding of the physiology of crop yield formation and its interaction with the environment will be a prerequisite to the effective implementation of these breeding strategies[4].

Source and sink are the core concept of crop physiology that has been widely accepted as the crucial process dominating yield formation [5].Generally, source is a material producer and exporter, whereas sink is a material importer and consumer.So far, it has been commonly accepted that assimilates transported to the grain in C3cereals have three major sources:(i)photosynthesis of the leaf and its sheath;(ii)pre-anthesis reserves stored in the stems, mainly the internode parenchyma; and (iii) ear (wheat and barley) or panicle(rice and oat) photosynthesis.However, the proportion in terms of the contribution of these source organs to grain filling remains imperfectly understood because of methodological limitations [6].

During grain filling, the mature leaf has long been considered as the main source of carbon (C), nitrogen (N), and other mineral nutrients like P and K[7-9].Photosynthetic efficiency of leaves is a determinant of yield potential and thus a priority target of crop breeding.For Super Rice breeding in China,emphasis is laid on the architecture of the top three leaves and panicle within the canopy, in order to select ideotypes with higher photosynthetic capability [10].The importance of leaf photosynthesis for crop yield was exemplified by the recent work of South et al.[11],who constructed a metabolic pathway in transgenic tobacco plants that more efficiently recapture the unproductive by-products of photosynthesis with less energy lost.Under field conditions,these transgenic plants were ～40%more productive than wild-type plants.

On the other hand,non-foliar organs such as stem,sheath,ear or panicle (mainly glumes, bracts and awns) also contain chloroplasts and have photosynthetic function, and their contribution to grain filling has been recognized for cereal crops not only under drought or other abiotic stresses but also under good agronomical conditions [12].Particularly, ear photosynthesis possesses stronger drought tolerance in comparison with the flag leaf,as is explained by the refixation of respired CO2in the developing seeds and the persistence of photosynthetic components[13].The contribution of non-leaf organs varied among crop species, cultivars, and growing conditions.In wheat and barley,the reported contributions of ears to grain filling differ widely, ranging from 10% to 76%[13-15].Therefore,selection for higher ear photosynthesis has been proposed as an avenue to increase yield potential and improve adaptation to abiotic stresses[13,16].Much has been revealed concerning the role of non-leaf organs in wheat and barley, however, there is little information regarding the assimilate contribution of sheath, internode, and bracts to grain filling for rice.

In nature there are two stable isotopes of carbon,with the heavier13C discriminating against the lighter12C during photosynthetic carbon fixation.The carbon isotope composition (δ13C) of plant organs reflects its photosynthetic performance, stomatal conductance, and transpiration efficiency[17,18], and thus was used as a time-integrated physiological trait by crop breeders [19].In addition, compared with traditional intrusive approaches based on a differential prevention of photosynthesis in some parts of the plant such as by shading, application of herbicides, or simply defoliating leaf blades [20], use of δ13C may help to quantify the relative contribution of different photosynthetic organs in a non?destructive manner to avoid unwanted compensatory effects triggered by intrusive methods[6].

In this study, using samples of two-year field and pot experiments, we first measured the carbon isotope composition of different organs and revealed their spatial and temporal variation.Then we compared the δ13C in its natural abundance in the water-soluble fractions of the sheath,internode, and bracts with the δ13C in mature grains, in order to assess the relative contribution of these organs to grain filling.The objective of this study is to explain the actual significance of non-foliar organs for rice yield formation,with the hope to extend the basic knowledge of source and sink relations.

2.Materials and methods

2.1.Field and pot experiments

The field and pot experiments were conducted at Danyang Experimental Station, Jiangsu province, China (31°54′31″N,119°28′21″E).The soil type of the field experiment was clay loam, as reported by Wang et al.[9].Soil for pot experiments was taken from the same paddy field, weighing nearly 20 kg per pot.Total nitrogen (N) was 240 kg ha?1for the field experiment, and 2 g N pot?1for the pot experiment.Two fertilization modes with different base/topdressing ratios were applied: (1) N5-5: base/topdressing, 5/5; (2) N10-0: base/topdressing, 10/0.Topdressing was performed at the panicle initiation stage.

In 2018, 24 japonica rice cultivars were planted in the paddy field, focusing on the δ13C composition in bracts, mainly the lemma and palea (Table S1).Seeds were sown in the nursery beds on May 21th.Seedlings of 4-week old plants were transplanted into the paddy field, with two seedlings per hill.The plot area was 8.75 m2, 3.5 m in length and 2.5 m in width.Three hundred panicles with similar flowering date were selected and tagged.Pooled samples of the top three leaves and bracts were collected at 20 days after anthesis (20 DAA) and the panicles were collected at maturity (60 DAA).The panicle was subdivided into four parts: (1) superior grain (SG), from top primary and middle primary branches; (2) inferior grain (IG), from middle secondary and bottom secondary branches; (3) superior bracts (SB) of SG; and (4) inferior bracts (IB) of IG.

According to the results of 2018, Ningjing 8 was selected for the pot experiment in 2019, mainly because of its δ13C compositions in bracts, leaves and grains, as explained blow (Section 2.3).Seeds were sown in the nursery beds on May 26th.Seedlings of 4-week old plants were transplanted into the pot, with two seedlings per hill.The plastic pot was 30 cm in height and 34 cm in diameter.Each pot grew six hills.Whole plants were sampled at 21, 14, and 7 days before anthesis (DBA), and 0, 10, 20, 30, 40, 50, and 60 DAA.Plants were dissected into subsamples of leaf, sheath, internode, bracts, and grains.

Samples were heated at 105 °C for 1 h in an oven to deactivate the enzymes and then were dried to constant weight at 70 °C for 48 h.After that, dry matter was weighed and samples were milled into powder and stored in plastic bags at room temperature for isotope analysis.

2.2.Carbon isotope analysis

The water-soluble fraction (WSF) of the leaf, sheath, internode, and bracts was extracted according to the method of Yousfi et al.[21].In brief, 50 mg of leaf, sheath, internode or bract powder was suspended in 1 mL of Milli-Q water in a centrifuge tube for 20 min at 5 °C.After centrifugation at 12,000×g for 5 min at 5 °C, the supernatant containing the WSF was heated at 100 °C for 3 min to precipitate the heatdenatured proteins.After cooling, it was centrifuged again (12,000×g for 5 min at 5 °C) to remove the denatured proteins.An aliquot of 100 μL of supernatant containing the proteinfree WSF was transferred to tin capsules for carbon isotope analysis.About 5 mg of the dry matter (DM) of mature grains was used for carbon isotope analysis.

The stable carbon isotope composition (δ13C) in the DM and WSF samples was analyzed by an elemental analyzer coupled with an isotope ratio mass spectrometer (Delta C IRMS, ThermoFinnigan, Bremen, Germany).It was operated in continuous flow mode in order to determine the stable carbon (13C/12C) isotope ratios of the same sample.The13C/12C ratio of each organ was expressed in δ notation [22]: δ13C = (13C/12C)sample/(13C/12C)standard? 1, where ‘sample’ refers to plant material and ‘standard’ to international secondary standards of known13C/12C ratios calibrated against Vienna Pee Dee Belemnite calcium carbonate (VPDB), with an analytical precision of 0.10‰ (standard deviation).

2.3.Relative contribution of leaf, sheath, internodes, and bracts to grain filling

According to methodology proposed by Sanchez-Bragado et al.[23], the relative contribution to grain filling of different organs was estimated through a comparison of the WSF δ13C of these organs with the DM δ13C in mature grains.There are some assumptions for this approach.It considers the grain as the only sink organ,whereas the leaf,sheath,internodes,and bracts are the source organs.Another assumption is to neglect the δ13C fractionation due to translocation of assimilates from either the leaf, sheath, internode, or the bracts to the grain.Thus it is expected that the δ13C of the grain would directly reflect the isotopic signal resulting from the combinations of the δ13C of assimilates coming from the four organs.In addition, for this method to work, it has to assume that the δ13C of the grains be intermediate between the source organs.In the current study, the δ13C in grains of the cultivar used(Ningjing 8) is consistently between the lowest value of the leaf and other organs including bracts and internode, suggesting the applicability of the method of Sanchez-Bragado et al.[23].Therefore, the leaf is chosen as a reference and the relative contribution of bracts and internodes is calculated separately as follows:

where “a” is the relative contribution to grain filling, δ13Cgrainthe carbon isotopic composition of mature grain(on DM base),δ13Csourcethe carbon isotopic composition in the WSF of bracts or internode, and δ13Cleafthe carbon isotopic composition in the WSF of leaf blade.

2.4.Statistical analysis

Data were subjected to one-way analyses of variance(ANOVA) using the general linear model in order to calculate the effects of timing and nitrogen on the studied parameters.Samples were analyzed in triplicate and mean values were used for comparison.The analysis of variance was performed using Least Significant Difference(LSD)test in SPSS 17.0 (Statistical Product and Service Solutions,IBM).

3.Results

3.1.Dry matter accumulation

At whole plant level, DM accumulation increased progressively after 21 DBA, with the maximum rate at 30 DAA when most of the superior grains completed filling and the inferior grains were at their highest rate of assimilate accumulation(Figs.1 and 2).

For individual organs, DM of leaves peaked at anthesis,then decreased slowly to 50 DAA,and increased slightly until maturity.DM of sheath had a similar changing pattern with that of leaves,but it decreased more sharply at 20 DAA(Fig.1).Obviously, the variation of DM was largest for the internode,increasing until 10 DAA, and then decreasing to its lowest point between 20 and 30 DAA.Therefore,the changing pattern of the vegetative organs depended mainly on that of the internode.

Fig.1–Dry matter accumulation of leaf,sheath,internode and panicle.N5-5 and N10-0,N fertilization treatments with the base to topdressing ratio of 5:5 and 10:0,respectively.Samples of the pot experiment in 2019 are used.The naming of leaf,sheath is presented in reverse order,with flag leaf defined as leaf 1.On the contrary,the naming of internode is in its forward order,where internode 1 is the first elongation internode aboveground.Samples were analyzed in three biological replicates.Mean values with different superscript letters are significantly different according to Tukey’s LSD test(P＜0.05).

Fig.2– Dry matter accumulation of grains(A)and bracts(B).N5-5 and N10-0,N fertilization treatments with the base to topdressing ratio of 5:5 and 10:0,respectively.IG and SG,inferior and superior grains.IB and SB,bracts of inferior and superior grains.Samples were analyzed in three biological replicates.Samples of the pot experiment in 2019 are used.

The reproductive organs of the grain and bracts showed a similar pattern in terms of DM accumulation, increasing progressively since the beginning of grain filling and finally plateauing at maturity.Grains of superior and inferior spikelets completed the filling process at 30 and 40 DAA,respectively (Fig.2A).By contrast, bracts of superior and inferior spikelets basically stopped DM accumulation at 20 DAA(Fig.2B).

N fertilization mode had significant influence on DM accumulation.The N5-5 treatment that applied N topdressing at panicle initiation resulted in heavier grains (Fig.2), thus increasing grain yield per pot (data not shown).It exhibited a more drastic reduction of DM in the internode at 20 DAA than the N10-0 did(Fig.1),and this may be explained by the strong nutrient demand of grains especially the superior grains at that time.In general,grain weight of N5-5 was larger than that of N10-0 both for the superior and inferior spikelets.Conversely,weight of bracts for N5-5 was consistently lower than that of N10-0 for both kinds of spikelets.

3.2.Temporal and spatial variations of carbon isotope signature

There was a constitutive difference in δ13C among different organs(Fig.3),showing a consisitant trend in the order of leaf(?27.84‰) ＜ grain (?27.82‰) ＜ sheath (?27.24‰) ＜ bracts(?26.81‰) ＜internode (?25.67‰).Notably, δ13C of the sink organ, grain, was in the middle between that of the leaf and other source organs, indicating the possibility of calculating the relative contribution of different source organs by the method proposed by Sanchez-Bragado et al.[23].In addition,N topdressing lowered the δ13C in the organs examined(Table 1).

3.2.1.Leaf

δ13C of the leaf had a peak(the most negative)at 10 DAA.After that, it was reduced slightly until maturity (Fig.3).For different positions of leaf within the plant, δ13C presented a diminishing trend from the flag leaf (leaf 1; ?27.21‰) to the last leaf(leaf 7;?28.00‰)of the bottom plant(Fig.4A and B).N topdressing lowered leaf δ13C by 0.57‰ on average, with leaf 6 being the most responsive (Table 1；Fig.4).

Fig.3–Carbon isotope composition of leaf,sheath,internode,bracts and grain.N5-5 and N10-0,N fertilization treatments with the base to topdressing ratio of 5:5 and 10:0,respectively.Data are averaged across subsamples.For example,that of leaf is the mean value of all leaves from leaf 1 to leaf 7.Mean values with different superscript letters are significantly different according to Tukey’s LSD test(P＜0.05).

3.2.2.Sheath

3.2.3.Internode

δ13C of the internode rose gradually after 21 DBA and reached its maximum at 20 DAA, 10 days later than that of leaf and sheath (Fig.3).This time point of 20 DAA corresponds to the time when DM of the internode was at the lowest value, implying the substantial translocation of assimilates in the internode to the growing grains.For spatial variation, δ13C was highest for the first internode (?26.37‰), and there was no significant different among other internodes (Fig.4 E and F).N topdressing lowered internode δ13C by 0.37‰ on average, with internode 1 being more responsive (Table 1; Fig.4).

3.2.4.Bracts

Notably, δ13C of the bracts presented an obvious transition at the time point of anthesis, after which it increased drastically (more negative; Fig.3).This pattern indicates that the bracts may be the source organ of the photosynthetic assimilates.As shown in Fig.4G and H, δ13C of the superior bracts (?26.91‰) was lower than that of the inferior bracts (?26.71‰).N topdressing lowered δ13C by 0.04‰ and 0.05‰ for the superior and inferior bracts (Table 1).

3.2.5.Grain

δ13C of the grain increased after 10 DAA and peaked at 40 DAA (?28.01‰) when the grain filling process was almost completed (Fig.3).After 40 DAA it diminished gradually until maturity.Fig.4I and J shows that δ13C of the superior grains (?27.88‰) was lower than that of the inferior (?27.76‰).N topdressing reduced δ13C by 0.22‰, particularly in inferior grains (Tables 1, S1; Fig.4).

Table 1–Variations of carbon isotope composition (δ13C) among leaf, sheath, internode, bracts and grain and between nitrogen fertilization modes.

3.3.Relative contribution of different source organs to grain filling

Mathematically, the prerequisite of the estimation method proposed by Sanchez-Bragado et al.[23]is that δ13C in the DM of the grain falls into the middle between that of the WSF in the leaf and other source organs.It implies that the same slope and origin to zero need to be found between the δ13C of grains and the combined δ13C of the leaf and non?foliar organs like bracts, internode or sheath.To verify the feasibility of this method, we used the data of 2018 to estimate the relative contribution of bracts to grain filling in comparison with the leaf.

The relative contribution of the δ13C of the leaf (δ13CL) and those of the superior (δ13CSB) and inferior (δ13CIB) bracts that accounted for the δ13C of the grain (δ13CG) was assessed through a linear fit.The δ13CGwas used as a dependent variable and a combination of the δ13C in the WSF of leaf and bracts were used as the independent variables, with assignment of a different weight for the leaf and bracts depending on the water status accounted for by the δ13CG(Fig.5).Thus, the δ13C of superior bracts had a relative contribution of 60% (δ13CSB×0.60) and leaf 40% (δ13CL×0.40) towards the δ13CGthat ranged between?27.2‰ and ?27.8‰.Conversely, the relative contribution of the bracts was 73%and the leaf 27%when δ13CGvaried between?26.0‰ and ?26.6‰.In such a way, a linear fit was achieved close to the line with a slope of one and origin to zero for superior bracts (Fig.5A).A similar trend was observed for inferior bracts (Fig.5B), indicating the feasibility of this estimation method proposed by Sanchez-Bragado et al.[23].

Who has been drinking my water? said he; and the little hare gave a jump, and, pointing to the rabbit, he answered, Look there! it must be he! Why, there is mud all over his face and paws! The rabbit, frightened out of his wits, tried to deny the fact, exclaiming, Oh, no, indeed I never did; but Big Lion would not listen, and commanded them to cane23 him with a birch rod

When examining the data of 2019, only the δ13C of the internode and bracts at 10 DAA and 20 DAA agreed with the assumption of the estimation method used in the current study.Thus we calculated the relative contribution of internode, and bracts, with leaf as the reference.As shown in Table 2, the relative contribution of internode was as high as 32.64% and 42.56% at 10 DAA and 20 DAA, respectively.N topdressing lowered its contribution.Bracts presented a larger contribution than the internode, with superior bracts being higher than inferior bracts.Similarly, N topdressing reduced their contribution for both types of bracts.It should be pointed out the value of bracts was lower in 2019 than that in 2018, and this explain the difference in water status as indicated by δ13CGbetween the two experimental years (Table 1; Fig.5).

4.Discussion

4.1.Constitutive nature of δ13C in different organs

The carbon isotope composition(δ13C)is the result of multiple processes of physical, chemical and biological fractionation along with the carbon flow from leaf photosynthesis to grain filling.The δ13C of different photosynthetic organs thus have a constitutive nature[24].For example,it is evident that there is a general trend towards an enrichment in13C in certain organs such as seeds, in comparison with leaves.There are several possible explanations for this.One possibility is that fractionation may occur during export, phloem loading and unloading and transport of carbohydrates from the leaf to grains.In that sense, Brugnoli and Farquhar [25]found differences in discrimination between carbohydrates stored in rice internodes (as sucrose) compared to primary leaves,which could be attributed to fractionation during enzymatic reactions or transport processes.Similarly, the current study finds the significant difference in δ13C among leaf, sheath,internode, bracts, and grains.Another explanation is that some plant organs may be formed at different times and developmental stages,when the contribution of diffusion and carboxylation limitation may be significantly different.For example, grains are formed during the late stage of the life cycle when growing conditions are always more unfavorable because of decreased soil water availability.Hence,the carbon fixed at this time would be isotopically heavier, with a less negative δ13C.This might explain the spatial effect of leaf and sheath position within the plant on δ13C in individual leaf and sheath, with the upper leaf or sheath that develops lately having the lower δ13C value (less negative) than the bottom ones.Conversely,Yoneyama et al.[26]found no evidence for a large discrimination in assimilates after photosynthesis during transport in wheat, thereby stating there was no significant difference between the leaf and stem.Therefore,the constitutive nature of δ13C of different organs varied with crop species, and care should be taken when applying it as a physiological trait in crop breeding.

4.2.Contribution of bracts to grain filling

Traditionally, leaves were viewed as the main source organ supplying photosynthetic assimilates for grain filling.Improving leaf photosynthetic capacity has been the most important target for crop breeding, as exemplified by plant ideotypes that focus on radiation use efficiency by modifying the plant architecture for leaves to harness more light [27].However, growing evidence shows that other organs especially the ear or panicle play comparable roles in grain filling13C labeling experiments on durum wheat showed that during the beginning of grain filling the C fixed in the flag leaf was stored as structural C compounds,starch,and soluble sugars,and then respired,with only a small amount of soluble sugars translocated to the ear.On the other hand,the C synthesized in the ear was used for grain filling [15].In review of the literature available, ear photosynthetic assimilation by the awns and glumes account for 10% to 70% of dry matter accumulation in wheat grain, whereas the flag leaf blade contributes on average only 8%of grain C depending on water status [13,28].Although striking, these findings collectively suggest that the ear or panicle photosynthesis are as important as the leaf in yield formation.

Morphologically, rice reproductive organs are different from wheat.The main photosynthetic structures of rice panicle are lemma and palea, while that of wheat ear are glumes and awns.Compared with wheat, there are few reports of the contribution of the panicle to rice grain filling.Our previous study revealed a substantial contribution of protein-N from bracts to rice grains, being as high as nearly 10% [8].Further, pot experimental results assigned a considerable contribution of bracts to grain for N,Mg,and Zn,being 5.96%, 12.56%, and 12.34%, respectively, indicative of positive role of bracts in N, Mg, and Zn remobilization towards grains[9].In the current study, by comparing the carbon isotope compositions of leaf, bracts, and grain, we estimate the relative contribution of bracts and assign it is as high as 58.65%and 52.49%for superior and inferior grains at 20 DAA,respectively.In addition, δ13C of the bracts presented a quantum leap between anthesis and 10 DAA, implying a role in transition from sink to source organ.To our knowledge,this is the first report of the estimation of panicle contribution in rice by stable isotopic analysis.In overall terms, our findings together with related results on wheat strongly argue that the morphological trait of panicle should be deliberately considered in rice breeding programmes, in order to harness more light energy to fill the proximal grains.

4.3.Contribution of the internode to grain filling

Stem,consisting of nodes and internodes,is not only structural support, but also a transitional storage pool of C assimilates.The buffering effect of stem nonstructural carbohydrates(NSCs)on sink strength constitutes a facet of physiology that is fundamental to all plant life [29,30].As uncovered by14C technique, 68% of the accumulated carbohydrate in stem was translocated into the grain, 20% was respired during the ripening period,and 12%remained in the vegetative parts.For the carbohydrate translocated,it was equal to about 21%of the grain carbohydrate[31].In this study,we calculated the relative contribution of internode in comparison with the leaf, and found it was as high as 32.64% and 42.56% at 10 DAA and 20 DAA, respectively.Collectively, these findings suggest a substantial role of stem in grain filling.

In general, dry matter weight of stem shows a dynamic pattern of three phases:(1)increasing before anthesis to store photosynthetic assimilates from the leaf; (2) deceasing after anthesis to feed the grains; and (3) increasing near maturity because of the reduced requirement of the grains that filled almost completely [32].The present study shows a typical pattern of this change for the dry matter in internodes, the main part of stem.Interestingly, the stem kept increasing its weight until 10 DAA, implying that it has dual roles as both source and sink organs during the early stage of grain filling.Further,a sharp decrease was observed at 20 DAA and 30 DAA,and this can be explained by the immediate requirement of grains whose filling process is at its high rate or intensity.The refilling of dry matter at late stage may be interpreted as the result of the surplus leaf photosynthesis that exceeds the sink demand of growing grains.According to these findings, the dry matter in stems can be viewed as an integrative index for source?sink balance at the whole plant level.

4.4.Carbon isotope signature as the physiological index for crop breeding

Physiological traits like canopy temperature, chlorophyll fluorescence, and carbon isotope composition were successfully used to breed elite cultivars with physiologically sound foundations [33].As an exemplar, carbon isotope discrimination(CID;Δ13C),the positive value of δ13C,provides an integrative assessment of leaf transpiration efficiency(TE)in C3species [24].It has been identified as a useful trait indicating TE, and some commercial varieties with drought resistance have been released based on selection for CID[34].

Nevertheless, some consideration has to be given when applying CID for crop breeding.Firstly, there are inherently temporal variations of δ13C in plant organs or tissues.As in this study, δ13C in leaf, sheath, internode,bracts, and grain is subject to change with development stage.It is only those at 10 DAA and 20 DAA that can be used for the estimation of organ contribution, and data from other stages do not meet the assumptions of our method.Secondly, δ13C is vulnerable to environmental cues, in particular water stress [28].In addition to N rate [23], our result indicates N fertilization mode also has a significant influence on δ13C in most of the organs measured, and the contribution of bracts and internode was reduced by N topdressing.The full details of this observation await further investigation.Lastly, the relation between CID and grain yield depends on crop genotype and growing environment.In wheat, high CID is positively associated with grain yields under favorable environments.By contrast, it is suggested that selection of genotypes with low CID may help to improve productivity by increasing water use efficiency under drought [35].The high level of inconsistency observed in the relationship between CID and yield has been considered as the greatest challenge of application of CID in wheat breeding for greater agronomic water-use efficiency [36].For rice, little is known regarding the relationship between CID and grain yield.Regression analysis of the data generated in the present study revealed a negative relation between δ13C of bracts and grain yield.However, it was not significant for the two nitrogen treatments (Fig.S1).

In addition, the method adopted in the current study for evaluating the relative contribution of source organs to grain filling still has it limitations.In the field experiments of 2018,we used 24 cultivars for genotyping the δ13C compositions among japonica rice cultivars and thereby evaluating the relative contribution ratios of bracts to grain filling.However,we can not calculate the ratios for 17 of the cultivars (Table S1),for that the δ13C compositions in leaves,bracts,and grains did not fit the assumption of the formula we used.This can be explained by the large spatial and temporal variations in δ13C composition within plant organs during grain filling stage,as revealed by the pot experiment in 2019.Thus,the composition of δ13C in plant organs is rather a complex issue, and much has to be done in the future to clarify the physiological mechanism of C fractionation process at whole plant level and its relevance to grain yield formation.

5.Conclusions

The constitutive nature of δ13C in leaf, sheath, internode,bracts and grains was revealed,showing a markedly temporal and spatial variation.The range of δ13C of the grain was between that of the leaf and other source organs (the internode and bracts), making it feasible to calculate the relative contribution of different source organs by the method of Sanchez-Bragado et al.[23].A substantial role of the internode and bracts was uncovered by this calculation, as also confirmed by the growth analysis on the basis of dry matter accumulation and translocation.Although surprisingly, the internode and bracts have a contribution at least comparable to the leaf that is traditionally viewed as the main source organ.Our results indicate that δ13C analysis is applicable to studies on source and sink relationships, and can be used as an integrative methodology for rice breeding.However, care has to be taken when applying it, in particular the notable difference between crop species, the strong influence of growing environment, and considerable positional variations within plant and temporal variations across growth stages.In addition, mechanisms controlling C fractionation processes from the leaf to the grains still demands thorough investigation.

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2020.06.011.

Declaration of competing interest

Authors declare that there are no conflicts of interest.

Acknowledgments

The research was supported by the National Key Research and Development Program of China(2017YFD0300103),the National Natural Science Foundation of China (31771719), and National High Technology Research and Development Program of China(2014AA10A605).Rothamsted Research receives strategic funding from the Biological and Biotechnological Sciences Research Council of the United Kingdom.Matthew Paul acknowledges the Designing Future Wheat Strategic Program(BB/P016855/1).

Author contributions

Mengjiao Jiang had the main responsibility for data collection and analysis, Hongfa Xu, Nianfu Yang contributed to data collection and analysis,Ganghua Li and Matthew Paul revised the manuscript, and Yangfeng Ding and Zhenghui Liu (the corresponding authors) had the overall responsibility for experimental design, project management, and manuscript preparation.